Monitor Ceases to Receive Input In-Game; GPU Fan at 100%
9 replies, posted
I've had my 1080 TI for around two months now without any sort of issue, but in the last two weeks I've begun running into something nasty.
This has happened four times now, but it's frequent enough to have me worried. With no apparent trigger, the GPU fails to output to my monitor, screen shutting off and the GPU fans crank to full bore.
It took a few times for me to realize, but the system seems to arrest as well, as audio that had been playing also ceases. I don't know if the game crashes or if the PC outright bluescreened given the GPU blindness, but there's nothing I can do until I hard boot the system, upon which I receive proper output to the monitor instantly.
The problem isn't an instant crash or anything. I've had sessions perform normally for three to four hours before it happens again out of nowhere. And thus far this has ONLY occurred in conjunction with 3D gameplay, as my PC has remained on for days without me only to crash during gameplay later.
Temps don't seem like an issue. I don't have a live in-game readout or anything, but switching into Afterburner shows the game at a consistent 60C and the framerate (locked at 60fps by Rivatuner) never dips aside from loading in some modded shit for Fallout 4.
Interestingly enough though, rebooting and going right back into the game led to another such crash inside a half hour. After being given some time though, I had another four-hour session before it happened again.
Every bit of research I've put into this suggests an obvious card fault of some manner, but also the potential for a power issue. My PSU is certainly sufficient for my system... but it's also from goddamn 2010.
CORSAIR TX Series CMPSU
So it wouldn't shock me if after years of reliable service that this old thing was beginning to fail.
So I don't want to needlessly RMA the card only to find out the PSU was the problem... nor do I want to drop $150 (even if I was planning an upgrade soon anyway, but money got tight) for a new PSU and have the opposite happen.
Really I just wish I had a way to isolate the problem for sure.
I'm leaning towards a driver issue. Driver crashes normally cause these exact things.
Agreed, use Display Driver Uninstaller to completely strip all traces of the driver out in safe mode, then do a clean install of the latest drivers and see if it continues.
Tried it, safe mode and all. Full fresh install, complete with needing to reboot the damn computer to finish.
Didn't make it 20 minutes into Fallout 4.
Though the game doesn't seem to matter regardless of load. Anything using the 3D shader cores or whatever in the card will trigger. I tried to see if I could squeak by with some STALKER: Shadow of Chernobyl and it happened.
Removed 1080 ti for RMA and swapped for my 980. RMA when I wake up, but first I'm leaving Dark Souls 2 running at a bonfire overnight. No time to test with proper gameplay, but if the issue weren't with the card, this would pretty definitively spell that out. Though I WOULD like to do proper game sessions too to be certain.
Keep us posted on that, hopefully it was just the card.
Not an extensive test, but the 980 held fine overnight, and I was even able to clear a few areas and kill Demon of Song without incident.
Drove the 1080 ti to the MSI RMA processing center, because it's within 20 miles of me, and after explaining the issue nearly got handed a swapped card on the spot, but they didn't have another card in stock yet.
WAY better than mailing it in, let me tell you.
Will see what happens once it's shipped out.
I had this issue back in the day with my 8800 GTS and it was a faulty PCI-E slot. Had an SLI MoBo so I just used the other slot and all was well. If your new card gives you trouble, perhaps something to consider.
I know this probably doesn't help now, but I had almost the exact same issue with my GTX 1070. Ended up going down the same path of completing wiping and installing the drivers twice, under-clocking, monitoring temps, etc. Later on I opened up my case to discover the 6-pin connector was slightly out of place. After securing it in place all my problems vanished and it ran like a dream again.
Just based on my experience I want to lean towards saying it's a power issue somewhere down the line, but hopefully it was inconsequential like mine and your new card works perfectly.
If that IS the case, then I'm REALLY fucked, because my motherboard only has one pcie 3.0 x 16 slot, with a 2.0 equivalent beneath that. I don't use SLI, preferring a single gpu setup, so I didn't pay for a board with redundant slots.
If that happens, I'll be gimping my gpu hilariously until I can finally afford my upgrade.
Been planning to up to a system with a threadripper for all my 3d rendering hobbies, but that's postponed given my wife lost her job a month back and I've been paying all the bills. So until she's stable, all my money is stuck to essentials. I'm sure not hunting to pay for an lga 1150 mainboard, AND the price of another copy of Windows 7, let alone everything else to rebuild the same computer.
So... yeah, I'm really hoping it's the gpu as well.
Sorry, you need to Log In to post a reply to this thread.